Technical advances in the identification, cloning, expression, and manipulation of nucleic acid molecules and the deciphering of the human genome have greatly accelerated the discovery of novel therapeutics. Rapid nucleic acid sequencing techniques can now generate sequence information at unprecedented rates and, coupled with computational analyses, allow the assembly of overlapping sequences into partial and entire genomes and the identification of polypeptide-encoding regions. A comparison of a predicted amino acid sequence against a database compilation of known amino acid sequences allows one to determine the extent of homology to previously identified sequences and/or structural landmarks. The cloning and expression of a polypeptide-encoding region of a nucleic acid molecule provides a polypeptide product for structural and functional analyses. The manipulation of nucleic acid molecules and encoded polypeptides may confer advantageous properties on a product for use as a therapeutic.
In spite of the significant technical advances in genome research over the past decade, the potential for the development of novel therapeutics based on the human genome is still largely unrealized. Many genes encoding potentially beneficial polypeptide therapeutics or those encoding polypeptides, which may act as “targets” for therapeutic molecules, have still not been identified. Accordingly, it is an object of the invention to identify novel polypeptides, and nucleic acid molecules encoding the same, which have diagnostic or therapeutic benefit.
Bone morphogenetic protein (BMP) is a member of the transforming growth factor-beta family, which was originally identified as a factor promoting bone formation from a cartridge implant (Wozney et al., 1988, Science 242:1528-34; Celeste et al., 1990, Proc. Nat. Acad. Sci. USA 87:9843-47). BMP is also known to play an essential role during the early embryogenesis of the frog, the fly, and in mammals. The precise concentration of active BMP seems to be important for the specification of particular cell types (Dale et al., 1992, Development 115:573-85; Dosch et al., 1997, Development 124:2325-34). An activity gradient of BMP2/4 is observed in, for example, Xenopus embryos in which the lowest expression is detected at the dorsal tip and the highest expression at the ventral tip—establishing the dorsoventral axis determination in the embryo. In another example, the control of BMP concentration at specific sites of tissue development suggests a role for BMP in organogenesis. Control of BMP expression is achieved by either localized expression of the BMP gene products or through the influence of the BMP inhibitor chordin (CHD) (Sasai et al., 1994, Cell 79:779-90)—or short gastrulation (SOG) (Francois et al., 1994, Genes Dev. 8:2602-16).
CHD/SOG is a large secreted protein produced from the Spemann's organizer, the master-controlling region for the dorsoventral axis specification at the gastrulation stage of Xenopus embryogenesis. CHD/SOG functions as a dorsalization factor, as does Noggin (Smith and Harland, 1992, Cell 70:829-40), which is also secreted from the organizer. The Drosophila SOG has a transmembrane domain at its amino-terminus, suggesting that it may be a type II transmembrane protein (Francois et al., 1994, Genes Dev. 8:2602-16). It has been proposed that the carboxyl-terminal side (extracellular domain) of the Drosophila SOG is cleaved off. However, Xenopus CHD (Sasai et al., 1994, Cell 79:779-90), Zebrafish CHD (Schulte-Merker et al., 1997, Nature 387:862-63), and murine CHD (Pappano et al., 1998, Genomics 52:236-39) do not contain the transmembrane domain. Instead, these proteins have a signal peptide, and are secreted directly. The CHD/SOG polypeptide contains four repeats of the cysteine-rich domain (CR1-4) that is also found in a variety of extracellular matrix proteins such as collagen and thrombospondin.
CHD/SOG is known to bind to one of the ventralizing factors, BMP4 (Piccolo et al., 1996, Cell 86:589-98). BMP4 has been shown to be essential for embryonic development of posterior-ventral mesoderm in mice (Winnier et al., 1995, Genes Dev. 9:2105-16). The binding of CHD/SOG to BMP4 inhibits BMP4 activity by preventing BMP4 from binding to its receptor (Piccolo et al., 1996, Cell 86:589-98). In this respect, the functional relationship between CHD/SOG and BMP4 resembles that between OPG and OPGL, although CHD/SOG is not structurally related to the BMP receptors. The binding affinity of CHD/SOG to BMP4 is specific and tight (Kd=3×10−10 M (Piccolo et al., 1996, Cell 86:589-98), and seems to require proteolysis in order to effectuate the release of bound BMP4. This proteolysis is achieved by a specific metalloprotease—Tolloid (TLD) or BMP1—that cleaves CHD/SOG to liberate either, or both, the first (CR1) and last (CR4) CR motifs (Piccolo et al., 1997, Cell 91: 407-16). Whether or not CHD/SOG has other functions or an independent function through its own receptor remains to be determined.
One of the most important roles of CHD/SOG is to establish a BMP4 morphogen gradient (Jones and Smith, 1998, Dev. Biol. 194:12-17). BMP4 itself only migrates a short distance and seems to act essentially on the cell autonomously (Jones et al., 1996, Curr. Biol. 6:1468-75). In contrast, the BMP4 inhibitors Noggin and CHD/SOG appear to exert a long-range effect, thereby forming an activity gradient of BMP4.
BMPs also play important roles outside of early embryogenesis, for example in the organogenesis of lung, gut, kidney, skin, heart and teeth, as well as in the later stages of embryogenesis (Hogan, 1996, Genes Dev. 10:1580-94). Some BMPs are expressed in a very localized fashion while others are expressed widely in a tissue. The importance of the localized action of BMP for organogenesis has been supported by transgenic mouse experiments using constructs by which BMP concentration is artificially elevated throughout the target tissue. In the case of lung, BMP4 is expressed in the distal tips of epithelium in the developing lung, and when overexpressed with the surfactant protein C promoter, the development of a small lung in which the structural organization (i.e., branching) has been severely disrupted is observed (Bellusci et al., 1996, Development 122:1693-702). Since the putative BMP-activity gradient could also be disrupted by the transgene expression, BMPs expressed widely in the tissue could also play a role in the determination of the structural organization of a tissue.
Noggin is another BMP2/4 inhibitor secreted from Spemann's organizer (Zimmerman et al., 1996, Cell 86:599-606). The biological role of Noggin and its mode of action are similar to CHD/SOG in Xenopus. Although the most notable function of Noggin is, like CHD/SOG, dorsalization, Noggin null-mutant mice have shown a bone phenotype (hyperplasia of chondrocytes) instead of an early embryonic phenotype (McMahon et al., 1998, Genes Dev. 12:1438-52; Brunet et al., 1998, Science 280:1455-57). This suggests that CHL or even CHD might have a non-dispensable function in the later stage of embryogenesis.